Re: Which client is already supporting FIG?

Dave_Raggett (dsr@hplb.hpl.hp.com)
Wed, 5 Jan 94 10:18:41 GMT


Alejandro Rivero writes:

> Can anyone out there suggest a WWW client or HTML+ parser
> implementing FIG directive?

You may wish to read: ftp://15.254.100.100/pub/draft-raggett-www-html-00.ps

I am currently rewriting the draft and will publish a revised version soon.
The spec for FIG has changed slightly following the WWW/TEI meeting last
November at Cork/Ireland. The IMAGE element described in the above document
was dropped. The new version looks like:

<!-- figures which subsume the role of the earlier IMG element.

Behaves identically to IMG for align = top, middle or bottom.
Otherwise figure is inserted after next line break (soft or hard).
For align=left, the image is left aligned and text is flowed
on the right of the image, and similarly for align=right, with
no text flow for align=center (the default). The caption is
placed under the image.

The <A> element is used for shaped buttons handled by browser,
while the ISMAP mechanism sends pointer clicks/drags to server.
The text contained by this element is used for text-only displays
and authors should remember to provide effective descriptions,
including label text for shaped buttons.
-->
<!ELEMENT FIG - - (CAPTION?,(%text;)*)>
<!ATTLIST FIG
id ID #IMPLIED
align (top|middle|bottom|left|center|right) center -- position --
ismap (ismap) #IMPLIED -- server can handle mouse clicks/drags --
src %URL; #IMPLIED -- link to image data --
charset CDATA #IMPLIED -- eg "ISO-2022-JP" for japanese -->

<!ELEMENT CAPTION - - (%text;)+ -- table or figure caption -->
<!ATTLIST CAPTION
id ID #IMPLIED
charset CDATA #IMPLIED -- eg "ISO-2022-JP" for japanese -->

The FIG element is now formally allowed anywhere that IMG is permitted.
Shaped buttons are defined, as before, using the SHAPE attribute of the
<A> element to define shaped hypertext buttons, which can also be rendered
in the conventional way for text only displays. The spec includes code for
hit testing polygons specified by the shape attribute. I am working on an
HTML+ browser for X11/XLib which will be ready in early spring `94.

> I find it more powerful that the ismap atribute, as there are
> no interaction with the server, and the figt and figa tags have
> other important result: no source .gif is needed (except to
> define size and proportion of the box, if the FIGD mechanism is not
> implemented). If geografical maps are going to proliferate, it will
> be a really needed tag.

The FIGT/FIGA elements were dropped following the WWW Wizards Workshop
last year. People felt that the FIGT overlay mechanism needed more thought
and should be deferred. The FIGA shaped buttons were replaced by the SHAPE
attribute on the <A> element as explained above for greater consistency.

Dave Raggett